SEO claims board
87 SEO claims, tested on a live website
Each claim states what Google's documentation or an open web standard says. Each one was then checked on a real site with a named tool and given a verdict.
This board is a worked example from one engagement. The site isn't named: pages, people and vendors are removed, and large numbers are rounded. Data as of September 7, 2026. The verdicts measure how this one site conforms. They are not evidence that any fix changed rankings.
PassThe site does what the claim says.
GapIt doesn't, and the evidence shows where.
PartialPart of the claim holds, or it holds on only some pages.
N/ANothing on the site to test.
PendingNot decided yet. No evidence, no verdict.
CorrectedThe verdict changed, and the record stays.
Cluster 01 · 12 claims9 Pass · 2 Gap · 1 N/Arobots.txt and sitemap structure are sound. This cluster grew twice during testing, because verification proved the original list of 85 had holes: 1.8.1 (sitemap extension schemas) and 1.8.2 (a sitemap entry must name a live URL). Each new claim was written from the governing documentation, not from the defect that exposed the gap. Both carry the cluster's Gaps. 1.9 moved from Gap to Pass on re-measurement after a fix.
Open cluster 01 →
Cluster 02 · 9 claims7 Pass · 2 PartialCanonicalization is the strong part: Google's chosen canonical matches the site's declaration on all 25 priority pages, and every redirect is a server-side 301. The two Partials are 2.8 (soft-404 exposure, mostly fixed on re-measurement) and 2.9 (a small set of badly encoded URLs out of everything the sitemaps declare).
Open cluster 02 →
Cluster 03 · 8 claims6 Pass · 1 Gap · 1 PartialThe site delivers crawlers a prerendered document and visitors a client-rendered app. 3.1 moved from Gap to Pass after a fix, on the same tool and the same deciding field. Two things are carried openly rather than folded into the Pass: spinners elsewhere are reduced, not gone, and a few URLs outside the priority set still fail, in a variant our own check could not detect at first.
Open cluster 03 →
Cluster 04 · 7 claims5 Pass · 1 Gap · 1 PartialScope came first: the priority pages don't carry every markup type, so this cluster used a seeded, stratified sample of about 200 URLs across every sitemap class. Re-running the same sample after a fix moved no verdict, and that was the finding: the measurements were stable. Google's own tools, which had been blocked by a user-agent gap, could then confirm them.
Open cluster 04 →
Cluster 05 · 13 claims8 Pass · 2 Gap · 1 Partial · 1 N/A · 1 PendingThe site-side half is clean: HTTPS, redirects, HSTS, mixed content, resource crawlability and mobile/desktop parity all pass across about 180 URLs. The failures are all in field data, and the trend matters more than the level. Mobile LCP improved sharply, while desktop LCP, mobile INP and CLS all got worse. A single 28-day window would have shown none of that. 5.8 stays Pending: 12 of 25 pages gate content behind a click, and no vetted tool yet tests whether that content is absent from what Google indexes, so the verdict fails toward Pending.
Open cluster 05 →
Cluster 06 · 7 claims3 Pass · 2 Partial · 2 N/ATitles are in good shape. 6.5 is the story here: a Gap that turned out to be our own parser bug, then a Pass that was only true of the 25 priority pages, then a Partial once a purpose-built tool found 202 of 428 pages on boilerplate descriptions.
Open cluster 06 →
Cluster 07 · 6 claims2 Pass · 2 Partial · 1 N/A · 1 PendingHeading order is clean, with one h1 per page. The finding is 7.4: every page nests repeated navigation inside <main>, which is one template decision and one fix. Three tool defects were caught in piloting. One would have produced a false Gap across all 25 pages.
Open cluster 07 →
Cluster 08 · 8 claims1 Pass · 2 Gap · 1 Partial · 2 N/A · 2 PendingCrawl and rendered views agree on every verdict, so the findings are content gaps, not delivery problems. There is no authorship signal (8.4) and no citation apparatus (8.5) on any priority page, and 4 pages lack sameAs links (8.7). Judgement calls with no site-side observable are held rather than forced into a verdict.
Open cluster 08 →
Cluster 09 · 8 claims1 Pass · 1 Gap · 1 Partial · 1 N/A · 4 PendingTitles pass. Anchor text is the Gap: not one of the 25 pages passes 9.6. Alt text exists everywhere but is often reused across many images (9.7). Keyword repetition is reported without a verdict, because Google publishes no threshold and we won't invent one.
Open cluster 09 →
Cluster 10 · 9 claims9 PendingNot tested on this site yet. All nine claims are Pending. We haven't put a verdict on anything without evidence.
Open cluster 10 →